Did Bezos ever actually compare Amazon's work culture to Darwinism? The Times article quotes a former HR executive as saying that, not Bezos.
I've often seen newspaper articles that report XYZ, and then follow-up articles that comment on or extrapolate from XYZ without ever considering whether the original journalist got it right. I think that's what's happening here. The Times says that a former Amazon HR executive compares the company's work culture to Darwinism, and the author of this piece assumes that because the HR executive used that phrase, Bezos himself 'equate[s] “Darwinism” with “ruthless competition”' and then argues that Bezos is wrong. But the commentator has not considered whether his original assumption about Bezos' beliefs reflects reality.
I think that, while it's important not to put words in Bezos' mouth, altering that line in the article doesn't hurt the overall message, which is Bezos prefers harsh competition in the workplace, and that's counterproductive.
High performing people often prefer environments with high stakes and high risk/reward. Being harsh is awful for people who can't stand it, but empowering for those who thrive that way.
Imagine a situation where your job is CRUD development. You write PHP and CSS and JavaScript for a series of ad campaigns (say, every marvel movie for the next 10 years). It's repetitive drone work.
Then compare that to a situation where you have to send a rocket to space in the next 6 months or the company is out of business and you'll be homeless.
Define "better work"? Sounds like that second situation will lead to some crappy architecture with bug-ridden code. The rocket will likely fail and explode.
Meanwhile in the first scenario you've got breathing room to build some truly beautiful code that will pay dividends as the upfront cost is amortized over a decade of reuse. We see similar results in academia, where incredible research can be done because there is predictability and reliability in the environment in which it is being performed.
High-stress rushed environments just lead to shitty work/life balance and, at best, mediocre solutions that will provide a stop-gap solution until there is time to do something better.
Look at the incredible output from engineers at Google, Facebook, etc... Most tech companies do their best work once they've removed stress and timeline pressure from the equation. The trick is to get to that as early as possible in the company's life.
Either you are basically a human robot doing spec work or you can drive your own creative output for goals larger than yourself.
People don't know what they are capable of until they are pushed beyond limits they thought they had. The future isn't made by cautious people, it's made by being better than anybody thought could be accomplished.
"Create advertising units for campaign 3,401 of 16,032" isn't something where "beautiful" or "meaningful" comes into play. It's just drone work.
get to that as early as possible in the company's life.
We don't model startups on how a 30,000 employee company works. Completely different beasts.
Pretty sure the latter situation ends with homeless engineers and dead test pilots. Einstein developed his seminal theories working as a patent clerk; not sure you can argue without evidence that an insane, high-stress workload is conducive to anything other than burnout.
You do realize this is the website for 80-hour-work-week, launch-startup-in-3-months fans, right? This isn't the place for nonsensical "work life balance" (which really means "i'm lazy and don't want to do good work, i just want to waste everybody else's time").
No one's ever going to see this, but I feel like your reply merits a response.
I believe, simply put, that hours-per-week are as useful a metric of progress as SLOC. The fact is that for any nontrivial development effort, adding work hours does not noticeably improve time to delivery and, in fact, starts to degrade progress.
There are a few reasons for this. First, the most important effort in any significant project takes place before anyone starts slinging code. Design and strategy are obvious examples, but programming problems of sufficient complexity also don't benefit from throwing hours at them. For example, one of our current problems can be formalized as an adversarial non-Bernoulli bandit with an unknown budget. I don't think the architect conducting the literature review is going to produce better or faster work if he went from 50 to 80 hours a week.
Second, even when everyone's got their shovels out, there isn't a linear relationship between hours and project velocity. At some point the marginal productivity per hour plateaus and, after that, quality issues and technical debt start piling up. A team with the ability to intellectually come up for air is a team that is going to deliver a better strategy, a better design, a better experience and ultimately a better product. I'd rather take an extra six months to get to market than have to pivot half a dozen times (I say, as they box up the startup next door -- sure, they shipped fast, but their strategy and their product failed, and they found that you can't get your Series A subscribed to when you have zero market energy).
I do think that for some people there's an idea that work-life balance (and I refuse to use scare quotes) shows insufficient commitment or hustle. But would you want a surgeon who boasts about how fast he can cut a patient open, rather than his or her postoperative results? An engineer who can whip out a bridge design in half the time with only a 10% higher catastrophic failure rate? Are we so insecure in the intellectual value of what we do that we have to brag about how hard we work instead of the value we bring?
I disagree. What we do is difficult, and it is intellectually challenging. If a project demands that the dev team work 10, 14, even 16 hour days through the project timeline (and I've been in all three of those positions), then there's something wrong with the project, the taskings, or the stakeholders.
We are professionals. We are intellectuals. We are not measured by the sum of our hours, but the result of our efforts. Take pride in yourself, because when you're a manager -- if you're not one already -- you need to take pride in your team.
It's the difference between "established company" and "startup" though, isn't it?
Sure, at an established company with 50+ up to tens of thousands of employees, all these nuances and efficiency-by-working-less gains come into play.
But, when you are 3 people and you have to ship 6 features in the next 2 months, and each feature will require a minimum of 3,000 lines of code, and you can write 200 lines of code every 6 hours... your only option is to work faster and better and longer than any potential competition. Maybe you can pull two to two and a half 6 hour sessions every day instead of just one per day. There's always a race against the clock for runway.
Most of the "great" products we see from the past were created by driven teams overworking themselves to change the world. The entire 30/40 hour stable work with vacations and breathing room and weekends off comes later in the life of a company.
Perhaps, but the author is reflecting on the productivity of the entire group rather than on the individual. I cannot say whether the chicken story really is pertinent to corporate culture, but at a glance this analogy feels relevant.
Enjoying harshness for it's own sake is masochistic stupidity. What's the upside for the amazon opportunity? As far as I can read from what's reported, below market compensation for the reward of watching your fucking back every microsecond and possibility of...? Where do I sign up to get books and trinkets delivered to your door a day faster? Beats the shit, mission wise, out of trying to cure cancer...
The entire point of this article was that while it may be empowering for some individuals, the entire group output suffers, and as such is less effective than a system that empowers high performing groups instead. I mean, sure, it was taking a study and trying to overextend the result—but still, merely repeating that it's great for certain individuals misses the point entirely.
True. But there are a lot of reports from Amazon employees that they have a policy of culling the herd. Even Nick Hanauer, an original investor of Amazon, tells us that Bezos is pretty ruthless https://youtu.be/uIGeKMU9izo?t=10m56s
Whether or not the title is false, it's certainly clickbait, so we changed it to a representative sentence from the article, in accordance with the HN guidelines.
It's a great idea to regulate click-bait. I wonder if others find it to be click-bait though. The title represents the gist of the article. Click-bait would be a title far removed from the actual arguments or content in the article, IMO.
Can someone qualified comment on whether group selection is taken seriously among biologists now? I haven't followed the debate much but Dawkins at least has argued that group selection acts too slowly compared to individual selection to be an important force in evolution.
In either case I don't think it has much bearing on Amazon. In both cases evolution is just a metaphor/scienciness.
Dawkins introduced selection on the individual level in The Selfish Gene (great book, before he became all ranty and weird). This has pretty much killed group selection, but EO Wilson is still clinging on. I think his latest book was a bit too much cherry-picking in defense of group selection for some people.
24 comments
[ 2.9 ms ] story [ 50.3 ms ] threadI've often seen newspaper articles that report XYZ, and then follow-up articles that comment on or extrapolate from XYZ without ever considering whether the original journalist got it right. I think that's what's happening here. The Times says that a former Amazon HR executive compares the company's work culture to Darwinism, and the author of this piece assumes that because the HR executive used that phrase, Bezos himself 'equate[s] “Darwinism” with “ruthless competition”' and then argues that Bezos is wrong. But the commentator has not considered whether his original assumption about Bezos' beliefs reflects reality.
Is there any research to support that statement? Genuinely curious. Seems like it could be true, but also not.
Then compare that to a situation where you have to send a rocket to space in the next 6 months or the company is out of business and you'll be homeless.
Which situation motivates you to do better work?
Meanwhile in the first scenario you've got breathing room to build some truly beautiful code that will pay dividends as the upfront cost is amortized over a decade of reuse. We see similar results in academia, where incredible research can be done because there is predictability and reliability in the environment in which it is being performed.
High-stress rushed environments just lead to shitty work/life balance and, at best, mediocre solutions that will provide a stop-gap solution until there is time to do something better.
Look at the incredible output from engineers at Google, Facebook, etc... Most tech companies do their best work once they've removed stress and timeline pressure from the equation. The trick is to get to that as early as possible in the company's life.
Either you are basically a human robot doing spec work or you can drive your own creative output for goals larger than yourself.
People don't know what they are capable of until they are pushed beyond limits they thought they had. The future isn't made by cautious people, it's made by being better than anybody thought could be accomplished.
"Create advertising units for campaign 3,401 of 16,032" isn't something where "beautiful" or "meaningful" comes into play. It's just drone work.
get to that as early as possible in the company's life.
We don't model startups on how a 30,000 employee company works. Completely different beasts.
I believe, simply put, that hours-per-week are as useful a metric of progress as SLOC. The fact is that for any nontrivial development effort, adding work hours does not noticeably improve time to delivery and, in fact, starts to degrade progress.
There are a few reasons for this. First, the most important effort in any significant project takes place before anyone starts slinging code. Design and strategy are obvious examples, but programming problems of sufficient complexity also don't benefit from throwing hours at them. For example, one of our current problems can be formalized as an adversarial non-Bernoulli bandit with an unknown budget. I don't think the architect conducting the literature review is going to produce better or faster work if he went from 50 to 80 hours a week.
Second, even when everyone's got their shovels out, there isn't a linear relationship between hours and project velocity. At some point the marginal productivity per hour plateaus and, after that, quality issues and technical debt start piling up. A team with the ability to intellectually come up for air is a team that is going to deliver a better strategy, a better design, a better experience and ultimately a better product. I'd rather take an extra six months to get to market than have to pivot half a dozen times (I say, as they box up the startup next door -- sure, they shipped fast, but their strategy and their product failed, and they found that you can't get your Series A subscribed to when you have zero market energy).
I do think that for some people there's an idea that work-life balance (and I refuse to use scare quotes) shows insufficient commitment or hustle. But would you want a surgeon who boasts about how fast he can cut a patient open, rather than his or her postoperative results? An engineer who can whip out a bridge design in half the time with only a 10% higher catastrophic failure rate? Are we so insecure in the intellectual value of what we do that we have to brag about how hard we work instead of the value we bring?
I disagree. What we do is difficult, and it is intellectually challenging. If a project demands that the dev team work 10, 14, even 16 hour days through the project timeline (and I've been in all three of those positions), then there's something wrong with the project, the taskings, or the stakeholders.
We are professionals. We are intellectuals. We are not measured by the sum of our hours, but the result of our efforts. Take pride in yourself, because when you're a manager -- if you're not one already -- you need to take pride in your team.
Sure, at an established company with 50+ up to tens of thousands of employees, all these nuances and efficiency-by-working-less gains come into play.
But, when you are 3 people and you have to ship 6 features in the next 2 months, and each feature will require a minimum of 3,000 lines of code, and you can write 200 lines of code every 6 hours... your only option is to work faster and better and longer than any potential competition. Maybe you can pull two to two and a half 6 hour sessions every day instead of just one per day. There's always a race against the clock for runway.
Most of the "great" products we see from the past were created by driven teams overworking themselves to change the world. The entire 30/40 hour stable work with vacations and breathing room and weekends off comes later in the life of a company.
Jobs at public tech companies also come with large stock packages that become usable after just a few short years.
The entire point of this article was that while it may be empowering for some individuals, the entire group output suffers, and as such is less effective than a system that empowers high performing groups instead. I mean, sure, it was taking a study and trying to overextend the result—but still, merely repeating that it's great for certain individuals misses the point entirely.
https://news.ycombinator.com/newsguidelines.html
"X got Y all wrong!" is also an intrinsically baity construct.
If you don't think the title we gave it is representative, we're always open to better ones.
In either case I don't think it has much bearing on Amazon. In both cases evolution is just a metaphor/scienciness.
A few years back Steven Pinker wrote a thing about the "false allure" of group selection based on muddy definitions: http://edge.org/conversation/steven_pinker-the-false-allure-...
Jerry Coyne summarised the debate in NYRB: http://www.nytimes.com/2011/09/11/books/review/the-neighborh...